J. Wang, J. Zhou, J. Liu, P. Wonka, J. Ye
Advances in Neural Information Processing Systems 27, conference, pp. 1053-1061, (2014)
The`1-regularized logistic regression (or sparse logistic regression) is a widely used method for si- multaneous classication and feature selection. Although many recent e orts have been devoted to its ecient implementation, its application to high dimensional data still poses signicant challenges. In this paper, we present a fast and effective sparse logistic regressions creening rule (Slores) to identify the \0" components in the solution vector, which may lead to a substantial reduction in the number of features to be entered to the optimization. An appealing feature of Slores is that the data set needs to be scanned only once to run the screening and its computational cost is negligible compared to that of solving the sparse logistic regression problem. Moreover, Slores is independent of solvers for sparse logistic regression, thus Slores can be integrated with any existing solver to improve the eciency. We have evaluated Slores using high-dimensional data sets from di erent applications. Extensive experi- mental results demonstrate that Slores outperforms the existing state-of-the-art screening rules and the eciency of solving sparse logistic regression is improved by one magnitude in general.